How privacy-preserving technology can advance satellite collision detection

Tyler Mitchell By Tyler Mitchell Jun14,2024

Wider participation in the space economy brings both opportunities and challenges, including a heightened risk of satellite collisions. Since no one group controls space, the open sharing of space situational awareness (SSA) data is critical to ensure safe space operations. However, this data can be sensitive for governments and operators, so in order to facilitate secure, privacy-preserving collaboration, the space sector should look to the advanced technologies being leveraged in other fields.

An evolving data landscape

With over 8,000 governmental and private satellites in orbit, the number of countries, regions, and commercial entities with SSA capabilities has proliferated. One of the most notable examples is the Space Surveillance and Tracking program launched by the European Union in 2023. The commercial SSA sector, too, has developed rapidly and now has more sensors in the southern hemisphere than the United States government.

Satellite operators are increasingly integrating multiple data sources, presenting an opportunity to improve the resilience and redundancy of collision avoidance systems. Some countries, including Canada, Japan and Australia have defined interoperability with U.S. systems as a goal in developing their own SSA systems. In the U.S., the Office of Space Commerce is also focused on international cooperation in SSA.

The need for data sharing

It has been clear that the global problem of space debris can only be solved through cooperation since at least the Iridium-Cosmos crash in 2009, after which the U.S. government began making more SSA data publicly available and notifying all entities of collision risks.

With the proliferation of satellites, not to mention the increased security risks ushered in with advancements in AI, the need for secure data sharing processes is growing. Additionally, even with all the regulatory progress that has been made in recent years, it will take time to action these policies, therefore making it critical that governments and corporations work together to explore ways of mitigating these potential risks. 

Any data sharing arrangement inevitably involves a tradeoff between privacy and utility. Because SSA data is highly sensitive, the U.S. government has tended to emphasize privacy in data sharing, even going as far as adding synthetic noise to public domain SSA data. These privacy considerations can reduce the accuracy of the space object tracking information on which other parties rely. 

This has, at times, led to trust issues around SSA data. A 2021 paper published by the ESA’s Space Debris Office revealed a general dissatisfaction with the quality and timeliness of SSA data, which at the time was mostly provided by the U.S.

The Office of Space Commerce acknowledges the crucial need for alignment on standards and best practices for sharing SSA data, noting in a May 2024 report that this is “an important step in facilitating international coordination and ensuring clear and efficient services for spacecraft operators.”

AI for satellite collision avoidance

As the quantity of data and the complexity of the satellite environment grows, space traffic management is increasingly vested in leveraging AI. For example, the ESA already held the Collision Avoidance Challenge in 2019, using a dataset of real-world historical conjunction data messages released by the ESA to develop and train AI models.

When it comes to AI, the more (quality) data, the better the model. This means that the data of multiple, collaborating parties is more valuable than the data of any one public or private entity, and is likely to support more accurate collision avoidance models.

PETs enable cross industry collaboration 

Innovation in developing the space economy must be matched by innovation in keeping it safe and secure. When it comes to data collaboration, there is a broad scope for cross-disciplinary learning, as different sectors have historically grappled with similar challenges. 

Privacy-enhancing technologies (PETs) are uniquely positioned to enable organizations to collaborate on sensitive data. PETs use cryptographic techniques to safeguard data during computation, minimizing an information system’s possession of sensitive data without losing functionality. 

Because PETs embody fundamental data protection principles, they can help organizations comply with stringent data privacy regulations. For example, PETs can address the imperative to minimize personal data use under the GDPR. 

PETs encompass a range of technologies that all aim to protect sensitive data, including Secure Multiparty Computation (a cryptographic protocol that distributes a computation across multiple parties, where no individual party can see the other parties’ data) and Fully Homomorphic Encryption — an encryption scheme that enables analytical functions to be run directly on encrypted data while yielding the same encrypted results as if the functions were run on plaintext.

In addition to preserving data privacy, PETs also address security concerns, as encryption-in-use safeguards even against quantum attacks. While to date few publicly known cyberattacks have targeted space systems, multiple countries possess the capabilities to conduct such attacks, and a growing number of non-state actors are discovering vulnerabilities in commercial satellite systems. 

PETS in action

Some industries that are already beginning to leverage PETs for collaboration on sensitive data can provide relevant use cases for the space sector. 

One relevant case study comes from the financial sector. Data scientists are exploring options to train AI models that detect fraud, and are collaborating with financial partners to train those models on more extensive datasets. However, security and compliance considerations pose obstacles to collaboration. In partnership with Inpher, the data science team at the cross-border payment service provider BNY Mellon built an innovative, collaborative fraud detection framework that leverages Secure Multiparty Computation to enable computations without revealing partners’ data. This led to 20% better predictions of fraudulent transactions, and performance improvements available through the use of GPUs, which allow for increased data use and higher model precision.

In conclusion, in a context where advanced data analysis and collaboration on sensitive data are critical to maintaining space as a common good, the space sector should look to innovative technologies like PETs that enable privacy-preserving data collaboration in other industries.

Manuel Capel, a PETs evangelist, heads up Business Development within EMEA at Inpher, a company focused on cryptographic computing. Using his dual engineering and business background, Manuel works with clients to spearhead secure data collaborations while maintaining strict confidentiality of data inputs. Manuel has over two decades of experience at top tier consulting firms and Fortune 100 companies in data analytics and finance roles. 

Tyler Mitchell

By Tyler Mitchell

Tyler is a renowned journalist with years of experience covering a wide range of topics including politics, entertainment, and technology. His insightful analysis and compelling storytelling have made him a trusted source for breaking news and expert commentary.

Related Post

Leave a Reply

Your email address will not be published. Required fields are marked *